Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
7195112 | Reliability Engineering & System Safety | 2018 | 29 Pages |
Abstract
Research on risk assessment of the maintenance process for onshore oil and gas transmission pipelines has been attracting ever more attention from the academic community. Due to the existence of uncertainties, risk propagation can hardly be precisely and/or robustly assessed. Therefore, in this paper, considering that decision-makers prefer uncertainty-informed risk information rather than unreliable “precise” risk values, a new insight is provided to deal with risk assessment of the onshore pipeline maintenance process under uncertainty. The risk assessment model is built on the framework of quantitative risk assessment based on AHP and expert knowledge. Meanwhile, to represent and quantify uncertainty, interval analysis is utilized to extend the whole model into an interval environment. As a result, an interval quantified risk assessment model is established for the onshore pipeline maintenance process. The study shows that interval analysis can effectively internalize, represent, quantify and propagate the uncertainty in the risk assessment model. In the specific case of emergency maintenance for the Gangqing dual pipeline, the interval scores to respectively characterize the occurrence likelihood and consequence severity are computed. As a result, the uncertainty-informed overall risk of the emergency maintenance process is determined and intuitively pinpointed in an interval risk matrix. The risk rating of the case is estimated as Level 2, indicating that operations with respect to emergency maintenance are well organized and the possibility of accident occurrence is low. Thus, maintenance can be carried out well under supervision. Even if a secondary accident would occur, the accident scope will be quite small and emergency measures are adequate enough to control the development of the accident and reduce accident losses. Moreover, the sensitivity sorting of sub-indexes of occurrence likelihood is obtained as I11â¯>â¯I23â¯>â¯I13â¯>â¯I22â¯>â¯I34â¯>â¯I12â¯>â¯I33â¯>â¯I21â¯>â¯I31â¯>â¯I32, indicating that improvement in the management capacity (I11), normal operations (I13) and completeness of protection (I22) will effectively reduce the occurrence of accidents and improve operational safety. Furthermore, risk estimation under the condition of missing data is tackled by using Monte Carlo simulations and provides a reasonable option when crucial information is lacking.
Related Topics
Physical Sciences and Engineering
Engineering
Mechanical Engineering
Authors
Xuchao Yu, Wei Liang, Laibin Zhang, Genserik Reniers, Linlin Lu,